Financial Management

March 18, 2003

Office of the Inspector Generalof the

Department of Defense

DoD Payroll Withholding Data for FY 2002(D-2003-060)

A Regular Statement and Account of the Receipts and Expenditures of all public Money shall be published from time to time.

Article I, Section 9

Constitution ofthe United States

Report Documentation Page

Report Date 03 Mar 2003

Report Type N/A

Dates Covered (from... to) -

Title and Subtitle Financial Management: DoD Payroll Withholding Data for FY 2002

Contract Number

Grant Number

Program Element Number

Author(s) Project Number

Task Number

Work Unit Number

Performing Organization Name(s) and Address(es) OAIG-AUD (ATTN: AFTS Audit Suggestions) InspectorGeneral Department of Defense 400 Army Navy Drive (Room801) Arlington, VA 22202-2884

Performing Organization Report Number D-2003-060

Sponsoring/Monitoring Agency Name(s) and Address(es) Sponsor/Monitor’s Acronym(s)

Sponsor/Monitor’s Report Number(s)

Distribution/Availability Statement Approved for public release, distribution unlimited

Supplementary Notes

Abstract

Subject Terms

Report Classification unclassified

Classification of this page unclassified

Classification of Abstract unclassified

Limitation of Abstract UU

Number of Pages 32

Additional Copies To obtain additional copies of this report, visit the Website of the Inspector General of the Department of Defense at www.dodig.osd.mil/audit/reports or contact the Secondary Reports Distribution Unit of the Audit Followup and Technical Support Directorate at (703) 604-8937 (DSN 664-8937) or fax (703) 604-8932. Suggestions for Future Audits To suggest ideas for or to request future audits, contact the Audit Followup and Technical Support Directorate at (703) 604-8940 (DSN 664-8940) or fax (703) 604-8932. Ideas and requests can also be mailed to:

OAIG-AUD (ATTN: AFTS Audit Suggestions) Inspector General of the Department of Defense

400 Army Navy Drive (Room 801) Arlington, VA 22202-4704

Defense Hotline To report fraud, waste, or abuse, contact the Defense Hotline by calling (800) 424-9098; by sending an electronic message to Hotline@dodig.osd.mil; or by writing to the Defense Hotline, The Pentagon, Washington, DC 20301-1900. The identity of each writer and caller is fully protected.

Acronyms

APO Agency Payroll Office CSRS Civil Service Retirement System DFAS Defense Finance and Accounting Service FEGLI Federal Employee’s Group Life Insurance FERS Federal Employees Retirement System IG DoD Inspector General of the Department of Defense OPF Official Personnel File OPM Office of Personnel Management TSP Thrift Savings Plan RITS Retirement Insurance Transfer System

DISTRIBUTION: OFFICE OF PERSONNEL MANAGEMENT UNDER SECRETARY OF DEFENSE FOR PERSONNEL AND READINESS ASSISTANT SECRETARY OF THE AIR FORCE (FINANCIAL MANAGEMENT

AND COMPTROLLER) DIRECTOR, DEFENSE COMMISSARY AGENCY DIRECTOR, DEFENSE FINANCE AND ACCOUNTING SERVICE DIRECTOR, DEFENSE INFORMATION SYSTEMS AGENCY DIRECTOR, DEFENSE LOGISTICS AGENCY NAVAL INSPECTOR GENERAL AUDITOR GENERAL, DEPARTMENT OF THE ARMY DIRECTOR, CIVILIAN PERSONNEL MANAGEMENT SERVICES

Office of the Inspector General of the Department of Defense

Report No. D-2003-060 March 18, 2003 (Project No. D2002FH-0061)

DoD Payroll Withholding Data for FY 2002

Executive Summary

Who Should Read This Report and Why? This report is intended for use by the Inspector General, the Chief Financial Officer, and the Associate Director for Retirement and Insurance at the Office of Personnel Management. The report discusses the results of agreed-upon audit procedures developed by the Office of Personnel Management. Agreed-Upon Procedures. Office of Management and Budget Bulletin No. 01-02, October 16, 2000, requires all Federal agencies to review their civilian employees’ retirement, health benefits, and life insurance payroll withholdings. The Inspector General, Chief Financial Officer, and Associate Director for Retirement and Insurance of the Office of Personnel Management developed specific agreed-upon procedures to review civilian employees’ withholdings and are therefore responsible for the adequacy of the agreed-upon procedures. We applied the agreed-upon procedures in accordance with the standards established by the American Institute of Certified Public Accountants. Background. Defense Finance and Accounting Service pays about 682,000 employees with a total gross payroll of about $1.4 billion within a given pay period through 11 payroll offices. Results. The payroll withholding amounts that the Defense Finance and Accounting Service reported to the Office of Personnel Management did not exactly match the supporting detail they provided. This is a repeat issue from prior Inspector General of the Department of Defense audit reports, and the Defense Finance and Accounting Service is developing computer storage to resolve the issue. We also noted differences between the Office of Personnel Management and Defense Finance and Accounting Service SF 2812s for military deposit cash payments. Withholding Data Discrepancies. We selected a sample of 280 employees and compared their payroll withholdings to authorizations in their Official Personnel Files. This comparison revealed 34 discrepancies (percentages apply to the sample of 280 not to the whole population) as follows:

• 3 gross pay discrepancies (1.07 percent), • 13 life insurance withholding discrepancies (4.64 percent), • 5 health benefits withholding discrepancies (1.79 percent), • 9 Thrift Savings Plan discrepancies (3.21 percent), • 3 Federal Employees Retirement System withholding discrepancies

(1.07 percent), and • 1 Civil Service Retirement System withholding discrepancy (0.36 percent).

ii

Calculations Required by Agreed-Upon Procedures. We recalculated headcounts, life insurance, and health benefits for the payroll detail files. The differences between the totals we recalculated and the Defense Finance and Accounting Service payroll files were less than the reporting threshold criteria established in the agreed-upon procedures. Conclusion. We performed the agreed-upon procedures specifically pertaining to payroll. We were not engaged to and did not perform an audit with the objective of expressing an opinion on the withholdings and contributions for health benefits, life insurance, retirement, and on the employee headcount of DoD. Therefore, we are not expressing an opinion. However, we performed additional procedures based on generally accepted government auditing standards. In general, the Defense Finance and Accounting Service and supporting DoD organizations could improve internal controls over the accuracy of the payroll amounts withheld and remitted to the Office of Personnel Management. The withholding amounts determined by performing the agreed-upon procedures differed from the withholding amounts presented in Defense Finance and Accounting Service reports. The differences were less than the reporting threshold criteria established in the agreed-upon procedures. Management Comments and Audit Response. The Defense Finance and Accounting Service commented that it enacted a new procedure to prepare supplemental SF 2812s for any unreported military deposit cash payments. Resolution of future unreported military deposits cash payments will occur through a system change to the Defense Civilian Pay System. The management comments are partially responsive because they will affect future payments, but not past payments.

The Office of Personnel Management confirmed that the Defense Finance and Accounting Service has prepared supplemental SF2812s. However, the supplemental SF2812s do not correct the discrepancies for military deposit cash payments and service credit addressed in our draft report. An itemization of the discrepancies in military deposit cash payments is in Table 1 of this report. The Defense Finance and Accounting Service confirmed that it implemented the system change to resolve the military deposit cash payments, which we address in this final report.

In this final report, we recommend that the Defense Finance and Accounting Service report military deposits and service credits listed in Table 1 of this report to the Office of Personnel Management. We request that the Defense Finance and Accounting Service provide comments by May 16, 2003, to explain how and when it has reported military deposit cash payments and service credit as listed in Table 1 of this report. A discussion of the management comments is in the Independent Auditor’s Report section, and the text is in the Management Comments section.

Table of Contents Executive Summary i

Independent Auditor’s Report

Overview 1 Attachment of Agreed-Upon Procedures and Associated Findings 6

Appendixes

A. Scope and Methodology 17 Management Control Program Review 17 Prior Coverage 18

B. Report Distribution 20

Management Comments

Defense Finance and Accounting Service 23

1

Independent Auditor’s Report

Overview We have performed the employee payroll and withholding data procedures agreed upon by the Inspector General, the Chief Financial Officer, and the Associate Director for Retirement and Insurance, Office of Personnel Management (OPM). The sufficiency of these procedures is solely the responsibility of the three officials agreeing to them. Therefore, we make no representation regarding the sufficiency of the procedures described in the attachment. For FY 2002, the OMB guidance added an additional requirement to obtain management comments concerning this audit, thus requiring a report rather than a memorandum as in past years. We provided a draft report to OPM and their contactor, KPMG, in November 2002, which satisfied their initial requirement to test the reliability on the DoD transfer of payments. We will provide this report to OPM to satisfy the requirement for management comments. We applied the procedures to the employee withholdings and employer contributions reported on the Report of Withholdings and Contributions for Health Benefits, Life Insurance, and Retirement for the payroll periods ended October 20, 2001; January 26, 2002; February 23, 2002; and March 09, 2002. We also applied the agreed-upon procedures to the Supplemental Semiannual Headcount Report as of February 23, 2002. We applied these agreed-upon procedures in accordance with the standards established by the American Institute of Certified Public Accountants in the Statement of Auditing Standards No. 87, “General-Use and Restricted–Use Reports.” Comparison of Amounts Withheld and Remittance to OPM. In general, the Defense Finance and Accounting Service (DFAS) and supporting DoD organizations could improve internal controls over the accuracy of the payroll amounts withheld and remitted to OPM. The amounts determined by performing the agreed-upon procedures differed from the amounts presented in DFAS reports, and the differences were within the parameters prescribed in the agreed-upon procedures. In prior years, DFAS did not maintain a historical permanent record of the data summarized for transmission to OPM. However, DFAS has begun to record the data summarized for transmission to OPM in a permanent record called the SF 2812 file. A system change request drove this requirement to record the data summarized for transmission to OPM. The system change request referred to as “the data-to-data project” allows the mechanized transfer of employee data, including history data, between databases. This mingling of history data will negatively impact the ability of DFAS to provide accurate, detailed audit trail data to support the summary data reported to OPM. Once DFAS completes the data-to-data project, the Defense Civilian Pay System will not be able to provide the detailed data supporting the payroll withholding information submitted to OPM, and the electronic record referred to as the SF 2812 file will have to provide this information. Payroll File Totals. We totaled the payroll files that included about 681,599 employees within a given pay period, with a total gross payroll of about $5.5 billion for the 4 pay periods we reviewed. The payroll withholding amounts DFAS reported to OPM exceeded the footings (that is, totals) of the DFAS database by $0.533 million, for an overall error rate of 0.159 percent. This is not an improvement from FY 2001, when the payroll amounts DFAS reported to OPM exceeded our footings of the DFAS database by $0.018 million, for an overall error rate of 0.008 percent. The differences found this year

2

are greater than last year, but the differences are not material with respect to the DoD financial statements. However, considering the sensitivity of payroll, the differences, which range as high as .7887 percent, represent a material internal control weakness in the preparation and reporting of DoD payroll that should be addressed by management. DFAS maintains a database that can be retroactively adjusted, and the adjustments may create discrepancies between reported and adjusted amounts. Comparison of Payroll System Data to Official Personnel Files. We compared a sample of 280 employees from 8 payroll data files to documentary support for amounts paid and withheld as shown in Official Personnel Files (OPFs). This comparison resulted in 34 differences (percentages apply to the sample of 280 not to the whole population) between payroll system data and OPFs for:

• 3 gross pay discrepancies (1.07 percent), • 13 life insurance withholding discrepancies (4.64 percent), • 5 health insurance withholding discrepancies (1.79 percent), • 9 Thrift Savings Plan (TSP) discrepancies (3.21 percent), • 3 Federal Employees Retirement System (FERS) withholding discrepancies

(1.07 percent), and • 1 Civil Service Retirement System (CSRS) withholding discrepancy

(0.36 percent). Army Official Personnel Files. During our review of 139 Army OPFs, we identified 36 OPFs with potential discrepancies. We provided the Army with the names, social security numbers, and the nature of the potential discrepancies in each of the 36 files. The Army later provided us with forms printed from personnel file databases that explained inconsistencies between OPFs and the Defense Civilian Pay System data for 24 of the OPFs. We accepted personnel documents generated from personnel file databases that completely explained differences between data in 21 of the OPFs and data in the Defense Civilian Pay System. However, three of the OPFs still had unresolved inconsistencies with data in the Defense Civilian Pay System. After our initial review, we relied on forms printed by the Army for our use. Therefore, we reclassified the 21 OPFs with explained inconsistencies to “adequately supported and correct.” The Army sample still had 15 discrepancies out of the original OPFs with potential discrepancies even after considering the additional documentation. Causes of Discrepancies. Missing TSP-1 forms caused discrepancies in the sample. Other discrepancies were caused by miscalculations in the life insurance pay base. In addition, amounts withheld were inconsistent with documented elections for life insurance, health insurance, and TSP. The other withholding discrepancies were caused by failure to communicate payroll information, including Health Benefits Election forms, and delay in processing Life Insurance Election forms. We discuss the causes of the discrepancies in the attachment to this report.

Calculations Required. The agreed-upon procedures require us to compare the headcount of the employees using payroll data files with the headcount in the Supplemental Semiannual Headcount Report. There was less than a 1 percent difference between our headcounts of employees using payroll data files and the Supplemental Semiannual Headcount Reports. This was within the 2 percent reporting threshold allowed in the agreed-upon procedures.

3

Life Insurance. Our recalculations of Option A life insurance from the payroll data files supported the amounts reported to the OPM. The calculated amount of $1,285,172.98 was almost the same as the $1,283,751.61 reported to OPM. The difference between the amount calculated and the amount reported to the OPM ($1,421.37) did not exceed the 2 percent reporting threshold. We did not recalculate Option C because the data for multiple elections were not available.

Our recalculations of health insurance withholdings from the payroll data files supported the amounts reported to OPM. The amounts we calculated from the payroll data files only varied from the amounts reported to OPM by percentages between 0 and 2.14 percent in total, including employee withholding and agency contributions for each payroll file. This was much lower than the agreed-upon procedures reporting threshold of 5 percent for health insurance variances. However, in payroll file 380900 the employees’ withholding recalculated from the SF2812A exceeded the amount withheld by 5.26 percent for the pay period ended February 23, 2002.

Comparison of Total Amounts Transferred. We compared DFAS and OPM documentation for the total amounts transferred for the payroll periods sampled. We found that not all the amounts reported by the Defense Civilian Payroll System were the same as the amounts reported by the OPM Retirement and Insurance Transfer System. Specifically, the amounts reported for military deposit and service credit did not match in 5 comparisons. The payroll center files with military deposit and service credit not matching were the 380300 and 380500 payroll files located in Pensacola. Four of the DFAS 2812s compared exceeded the OPM 2812s by $10,770. In the other comparison, the OPM 2812 exceeded the DFAS 2812 by $1,943. Because of the mismatches discovered during the comparisons of OPM and DFAS 2812s, we subsequently obtained and compared the Form 2812s for payroll files 0300 and 0500 for 11 additional pay periods between October 6, 2001, and April 20, 2002. This test showed six additional comparisons that did not match. In total, we found 10 instances between October 6, 2001; and April 6, 2002; in which the DFAS 2812 exceeded the OPM 2812 by a total of $36,916. The dollar amount is important because the difference is concerned with cash or checks that could be subject to mishandling.

DFAS was aware of this deficiency in the Defense Civilian Payroll System prior to September 1, 2000. On September 1, 2000, personnel at the DFAS-Pensacola prepared a system change request to resolve this deficiency. DFAS completed the system change to systematically resolve military deposit cash payments in the Defense Civilian Pay System on December 1, 2002. Completion of the system change request allows the amounts reported to OPM by the Retirement Insurance Transfer system to be correct. However, DFAS-Pensacola took 2 years to complete this important system change request.

Management Comments. DFAS concurred and provided additional information. DFAS commented that it enacted a new procedure to prepare supplemental SF2812s for any unreported military deposit cash payments and that it has resolved this problem by implementing a system change in the Defense Civilian Pay System.

Audit Response. The Management Comments are partially responsive. OPM confirmed that DFAS has prepared supplemental SF2812s. However, the supplemental SF2812s do not correct the discrepancies for military deposit cash payments that we discussed in our draft report. We note 10 instances in which the DFAS 2812 exceeded the OPM 2812 by a total of $36,916. Table 1 summarizes the unreported military deposit cash payments.

4

Table 2 summarizes the supplemental SF 2812s filed by DFAS for payroll data files 97380300 and 97380500 for military deposits during FY 2002. We request that DFAS provide comments on the final report, to explain how and when it reported military deposit cash payments and service credit itemized in this report to OPM.

Table 1. Unreported Military Deposit Cash Payments During FY 2002

Pay Period Payroll Amount of Deposit Ended File Unreported October 6, 2001 97380500 $ 3,947 October 20, 2001 97380500 10,056 December 1, 2001 97380500 7,765 December 15, 2001 97380500 10,8781 December 15, 2001 97380500 122 January 26, 2002 97380300 178 January 26, 2002 97380500 456 March 9, 2002 97380300 80 April 6, 2002 97380300 743 April 6, 2002 97380500 2,688 Total $36,916 1Service Credit

Table 2. Supplemental SF 2812s Filed by DFAS During FY 2002 For Military Deposits

Pay Period Payroll Amount of Deposit Ended File Reported June 29, 2002 97380300 $11,615 June 29, 2002 97380300 21,220 Total $32,835

We performed the agreed-upon procedures specifically pertaining to payroll. We were not engaged to, and did not, perform an audit with the objective of expressing an opinion on the withholdings and contributions for health benefits, life insurance, retirement, and on the employee headcount of DoD. Therefore, we are not expressing an opinion. However, we performed additional procedures based on generally accepted government auditing standards. This report is intended for use by the Inspector General, the Chief Financial Officer, and the Associate Director for Retirement and Insurance, OPM. This report is prepared in the

5

format directed by Office of Management and Budget Bulletin No. 01-02, October 16, 2000, to address the results of the agreed-upon procedures. Accordingly, this report should not be used by those who have not agreed to the procedures and have not taken responsibility for the sufficiency of the procedures for their purposes. Appendix A discusses the scope and methodology used in accomplishing the agreed-upon procedures. Recommendation We recommend that the Director, Defense Finance and Accounting Service report military deposits and service credits listed in Table 1 of this report to the Office of Personnel Management.

6

ATTACHMENT OF AGREED-UPON PROCEDURES AND ASSOCIATED FINDINGS

This attachment contains the OPM agreed-upon procedures and the audit actions and results of accomplishing those procedures. Procedure. Obtain the Agency Payroll Office’s March Semiannual Headcount Report submitted to OPM and a summary of Retirement Insurance Transfer System (RITS) submissions for the current fiscal year. For retirement, health benefit, and life insurance, select any three RITS submissions for the current fiscal year, one of which coincides with the March Semiannual Headcount Report. Obtain payroll information for the periods covered by the RITS submissions selected. 1. Compare RITS submission data with payroll information by performing the following procedures: 1.a. Recalculate the mathematical accuracy of the payroll information. For cross-servicing agencies, if the internal controls are the same for all agencies serviced, it is only necessary to perform this procedure for one agency. Auditor Action. DFAS extracted all 11 of the payroll data files from the payroll history database and sent them to us by compact disc from its Pensacola, Florida, operating location. We totaled the 44 payroll data files (11 payroll files for 4 pay periods) with about $5.5 billion in total pay and about 682,000 employees in each payroll period. We totaled the life insurance withholdings, health insurance withholdings, CSRS withholdings, and FERS withholdings. The total withholdings calculated by audit from data provided by DFAS were $50.7 million for life insurance, $139.0 million for health insurance, $148.7 million for CSRS, and $22.7 million for FERS. Procedure. 1.b. Recalculate the mathematical accuracy of each RITS submission for the payroll information selected in step 1.a. Auditor Action. We recalculated the mathematical accuracy of each RITS submission for the payroll information for the payroll periods ended October 20, 2001; January 26, 2002; February 23, 2002; and March 9, 2002. OPM provided copies of the RITS submissions that we used for recalculation of the mathematical accuracy. Procedure. 1.c. Compare the employee withholding information shown on the payroll information obtained in step 1.a. for retirement, health benefits, and life insurance (as adjusted for reconciling items) to the related amounts shown on the RITS submission for the corresponding period. Auditor Action. We compared the employee withholding totals to the related amounts shown on the RITS submission for retirement, health benefits, and life insurance, as evidenced by a Form 2812 OPM produced from the RITS database. The payroll data file totals for life insurance and FERS equaled the related amounts shown on the RITS submission for the corresponding period. The payroll data file totals for health benefits and CSRS did not equal the amounts reported to OPM in the RITS. The health benefits

7

amounts reported in the RITS ($129,329,098) exceeded the payroll data file totals for health benefits ($128,729,887) by $599,211, or 0.4655 percent. The payroll data files for the CSRS ($137,512,490.90) exceeded the amounts reported in the RITS for CSRS ($137,446,144.42) by $66,346.48, or .0482 percent. The total differences, percentage differences, and high/low percentage differences are shown in Table 3.

Table 3. Differences Between Payroll Data Files and RITS Data Submissions

Reported Type of to OPM Total of Percent Withholding (millions) Differences Difference High/Low Percent Life $ 46.8 $0 .0000 0.0000/ 0.0000 Health 129.3 599,211.02 .4655 0.7887/ 0.2948 CSRS 137.5 66,346.48 .0482 0.0799/ 0.0045 FERS 20.9 0 .0000 0.0000/ 0.0000 DFAS provided us with an electronic extract from its database. Table 3 shows the differences between the payroll data files and the RITS data submission. The differences found this year are greater than last year, but the differences are not material with respect to the DoD financial statements. However, the differences, which range as high as .7887 percent, represent a significant internal control weakness in the DoD payroll, considering the sensitivity of payroll. We will be reporting this internal control weakness to DFAS. Procedure. 2.a. Randomly select a total of 25 individuals who were on the payroll system for all three of the RITS submissions selected and meet all the following criteria:

• covered by the CSRS or the FERS; • enrolled in the Federal Employees Health Benefits Program; • covered by Basic Life Insurance; and • covered by at least one Federal Employees Group Life Insurance (FEGLI)

optional coverage (Option A, B, or C). Auditor Action. We randomly selected 25 individuals from each of eight payroll data files in the Department of Defense with more than 30,000 employees who were enrolled in Federal retirement, health benefits, and life insurance programs.

8

Procedure. 2.b. Obtain the following documents, either in electronic or hard copy format, from the OPF for each individual selected in step 2.a. Hard copies can be originals or certified copies.

• all Notifications of Personnel Actions (SF-50) covering the pay periods in the RITS submissions chosen;

• the Health Benefit Registration Form (SF-2809) covering the pay periods in the RITS submissions chosen (note: a new SF-2809 is needed only if an employee is changing health benefit plans; therefore, the form could be many years old.); and

• the Life Insurance Election Form (SF-2817) covering the pay periods in the RITS submission chosen (note: a new SF-2817 is needed only if an employee is changing life insurance coverage; therefore, the form could be many years old).

Auditor Action. We obtained Notifications of Personnel Actions (SF-50), Health Benefit Registration Forms (SF-2809), and Life Insurance Election Forms (SF-2817) covering the pay periods in the RITS submission chosen. Procedure. 2.c. Via the agency personnel office, request a report from Employee Express for any health benefit transactions in that system for the individuals selected in step 2.a. Compare date of transaction with date on the certified copy of the SF-2809 requested in step 2.b. Confirm that the health benefit information to be used in the step 2.g. covers the pay periods in the RITS submissions chosen. Auditor Action. We requested copies from the agency personnel office of any automated health benefits elections (SF-2809) that could explain differences between OPFs and the Defense Civilian Pay System. The Army provided copies of personnel documents from the Army Benefits Center, and we directly accessed the Personnel Automated Records Information System for Air Force personnel documents. Procedure. 2.d. Compare the base salary used for payroll purposes and upon which withholdings and contributions generally are based with the base salary reflected on the employees’ SF-50s. Report any differences. Auditor Action. We compared the base salary used for payroll purposes with the base salaries reflected on the employees’ SF-50s. Out of 280 whose files were sampled, the base salaries used for payroll purposes were not supported by base salaries reflected on the SF-50s of 3 employees. Procedure. 2.e. For retirement, compare the plan codes on the employees’ SF-50s to the plan codes used in the payroll system. Report any differences. Auditor Action. We compared the plan codes on the employees’ SF-50s to the plan codes used in the payroll system. We did not note any differences between the retirement plan codes on the employees’ SF-50s and the retirement plan codes used in the payroll system.

9

Procedure. 2.f. Calculate the retirement amounts to be withheld and contributed for the plan codes on the employees’ SF-50s, based upon the official withholding and contribution rates required by law. Compare to actual amounts withheld and contributed. Report any differences. Auditor Action. We calculated the retirement amounts to be withheld and contributed for the plan codes from the employees’ SF-50s, based upon the official withholding and contribution rates required. We compared the retirement amounts we calculated to actual amounts withheld and contributed. We found four differences, one was due to a calculation error omitting environmental pay in the FERS salary base. The other two were due to errors in gross pay, which then affected the FERS salary bases. These two others showed no gross pay in the Defense Civilian Pay System, although the personnel files indicated that the employees were active and not on leave without pay. The remaining discrepancy was due to the OPF containing the incorrect SF-50. Procedure. 2.g. For health benefits, compare the employee withholdings and agency contributions with the official subscription rates issued by OPM for the plans and options elected by the employees, as documented by a Health Benefits Registration Form SF-2809 in the employees’ OPF or Employee Express. Report any differences. Auditor Action. We obtained the official subscription rates for health benefits issued by OPM for all plans and options available to Federal employees. We compared the employee withholdings and agency contributions with the official subscription rates issued by OPM for the plans and options elected by the employees, as documented by a Health Benefits Registration Form SF-2809 in the employees’ OPF. We found five discrepancies between the payroll system database and health withholding data in the OPFs. For two of the discrepancies, the health plan code on the SF-2809 in the OPF did not match the health plan code in the Defense Civilian Pay System. The amounts of the discrepancies were $4.37 per pay period and $3.68 per pay period. Two additional discrepancies occurred because there was no withholding for health benefits, and the employee was not paid. The OPF contained no explanation for the lack of pay. The amounts of these discrepancies were $48.94 and $75.74. The fifth health withholding discrepancy was due to a completed Health Benefits Registration Form in the OPF but with no corresponding health code or withholding in the Defense Civilian Pay System. The SF-50 in the OPF applicable to the pay period audited stated that the employee was not eligible for health benefits. The amount of the fifth discrepancy was $31.61. Procedure. 2.h. For life insurance, confirm that Basic Life Insurance was elected by the employees, as documented by a Life Insurance Election Form (SF-2817), in his/her OPF. Report any differences. Auditor Action. We reviewed evidence for life insurance election by reviewing Life Insurance Election Forms (SF-2809) and SF-50s. All of the OPFs had the required Life Insurance Election Forms. However, if a civilian employee in the Department of Defense does not waive life insurance or elect a life insurance option, they are automatically treated as electing Basic Life Insurance coverage. This treatment is consistent with the U.S. Office of Personnel Management FEGLI Handbook as shown on the OPM website.

10

The Handbook states that, “As a Federal employee, you automatically have Basic insurance (unless you are in a category of employees specifically excluded by law or regulation), unless you waive it on the Life Insurance Election form (SF 2817).” Procedure. 2.i. Calculate the withholding and contribution amounts for Basic Life Insurance using the following: For employee withholdings: Round the employee’s annual base salary to the nearest thousand dollars and add $2,000. Divide this total by 1,000 and multiply by $0.155 (for Agency Payroll Offices [APOs] with biweekly pay periods) or $0.3358 (for APOs with monthly pay periods). For agency contributions: Divide the employee withholdings calculated above by two. Compare to actual amounts withheld and contributed. Report any differences. Auditor Action. We calculated the withholding and contribution amounts for Basic Life Insurance by rounding the employees’ annual base salaries to the nearest thousand dollars and adding $2,000. For Federal Wage System employees, we added environmental differential to the base salaries in determining wages eligible for life insurance. Two of the sample OPFs that had elected Basic Life Insurance had calculation errors in Basic Life Insurance. One of these was due to not including environmental differential in wages eligible for life insurance. The amounts in error were $.62 and $.62. Procedure. 2.j. Also, for life insurance, compare optional coverage elected, as documented by an SF-2817 in the employees’ OPF, with optional coverage documented in the payroll system. Report any differences. Auditor Action. We obtained SF-2817 documents directly from employees’ OPFs and electronic personnel data files. We obtained life insurance optional coverage data from the Defense Civilian Pay System. We compared optional life insurance coverage elected as documented on the SF-2817s with optional life insurance coverage as recorded in the Defense Civilian Pay System. We identified three instances in which employees with optional life insurance coverage had incorrect optional coverage recorded in the Defense Civilian Payroll System. The amounts in error were $20.50, $7.96, and $5.28. In some cases, the election was recorded incorrectly, and in other cases, it appears that the election was not communicated to DFAS. Procedure. 2.k. Calculate the withholding amounts for optional life insurance using the following:

• For Option A: Determine the employee’s age group using the age groups provided for Option A in the FEGLI Program Booklet. The withholding amount is the rate listed in the FEGLI Program Booklet for that age group. Compare to amount withheld. Report any differences.

• For Option B: Inspect the SF-2817 to determine the number of multiples chosen

for Option B. Determine the employee’s age group using the age groups provided for Option B in the FEGLI Program Booklet. Round the employee’s annual rate

11

of basic pay up to the next 1,000, divide by 1,000, and multiply by the rate for the age group. Multiply this amount by the number of multiples chosen. Compare to amount withheld. Report any differences.

• For Option C: Inspect the SF-2817 to determine the number of multiples chosen

for Option C. Determine the employee’s age group using the age groups provided for Option C in the FEGLI Program Booklet. Multiply the rate for the age group by the number of multiples chosen. Compare to the amount withheld. Report any differences.

Auditor Action. We calculated the amounts for optional life insurance. In addition to the errors noted under 2.h., 2.i., 2.j., and 2.k., we identified three optional life insurance errors. Specifically, these errors were for $4.34, $1.40, and $.46. The errors for $1.40 and $.46 were due to including swing shift pay in the life insurable pay. Two other life insurance errors were due to the official personnel files not being up to date, and three other life insurance errors were due to life insurance calculation errors. In total, we identified 13 life insurance errors in all categories. Procedure. 3. Randomly select a total of 10 employees who have no health benefits withholdings from the payroll information corresponding to the three RITS submissions selected for testing. Request SF-2809s covering the pay periods in the RITS submissions chosen, either in electronic or hard copy format, from the selected employees’ OPFs. Hard copies can be originals or certified copies. Via the agency personnel office, request a report from Employee Express for any health benefits transactions in that system for the individuals selected. Inspect the documentation to determine that health benefit coverage was not elected. This can be determined in the following ways:

• absence of an SF-2809 in the OPF and no election of coverage made through Employee Express;

• an SF-2809 in the OPF with Section E checked (indicating cancellation of coverage) and no later election of coverage through Employee Express; or

• cancellation of coverage through Employee Express and no later election of coverage with an SF-2809.

Report any exceptions. Auditor Action. We randomly selected 10 employees per payroll data file who had no health benefits withholdings from the payroll information corresponding to the RITS submissions selected for testing. We reviewed the OPFs and electronic personnel databases for SF-2809s. We inspected the documentation to determine if health benefit coverage was elected. Except for one instance, we found no indication of election of coverage either in OPFs or in electronic files for employees who had no health benefit withholding. In one instance, discussed under 2.g., a file had an SF-2809 with health coverage elected, but the SF-50 showed that the employee was not eligible for health benefits. The OPF did not have a new SF-50 showing eligibility until well after the date of our review. No deductions were made for health benefits in the Defense Civilian Pay System for this employee during the periods we reviewed. We treated this as an error because the OPF and the Defense Civilian Pay System were inconsistent. In addition, the

12

OPF was internally inconsistent because the SF-50 showed that the employee was not eligible for health benefits, but the SF-2809 in the OPF reflected an election by the employee for health coverage. Procedure. 4. Randomly select a total of 10 employees who have no life insurance withholdings from the payroll information corresponding to the three RITS submissions selected for testing. Request the SF-2817s covering the pay periods in the RITS submissions chosen, either in electronic or hard copy format, from the selected employees’ OPFs. Hard copies can be originals or certified copies. Inspect the SF-2817 to determine that the employee waived or canceled Basic Life Insurance coverage. Report any exceptions. Auditor Action. We randomly selected 10 employees per payroll data file who had no life insurance withholdings according to the Defense Civilian Pay System corresponding to the three RITS submissions selected for testing. We requested, obtained, and reviewed the SF-2817s covering the pay periods in the RITS submissions chosen. We inspected the SF-2817s in all instances when the coverage was waived. No exceptions were noted. Procedure. 5. Recalculate the headcount reflected on the Semiannual Headcount Report selected for testing above, as follows. 5.a. Obtain existing payroll information supporting the selected Supplemental Semiannual Headcount Report selected for testing above, as follows:

• Benefit category (see Semiannual Headcount Report), • Dollar amount of withholdings and contributions, • Number enrolled (deductions made/no deductions), • Central personnel data file code, and • Aggregate base salary.

5.b. Recalculate the Headcount reflected on the Semiannual Headcount Report. If an electronic file is not available, a suggested method of recalculating the headcount is as follows: (1) estimate the number of employees per payroll register page by counting the employees listed on several pages, (2) count the number of pages in the payroll register, and (3) multiply the number of employees per page by the number of pages or count (using a computer audit routine) the number of employees on the payroll data file for the period. 5.c. Compare the results of payroll information from step 5.a. with the calculated headcount from step 5.b. to information shown on the Semiannual Headcount Report. 5.d. Report any differences (e.g., gross rather than net) greater than 2 percent between the headcount reporting on the agency’s Semiannual Headcount Report and payroll information from step 5.a. and the calculated headcount from step 5.b. Auditor Action. We obtained the DFAS Supplemental Semiannual Headcount Reports (see Table 4) for the pay periods ended February 23, 2002, for Payroll Offices 200, 300, 400, and 800 and March 9, 2002, for Payroll Offices 100, 500, 600, and 900 and compared those counts to the payroll data files from DFAS-Pensacola for the same period.

13

Table 4. Comparison of Employee Headcounts

Employee Count Employee

Payroll Report DFAS Payroll Headcount Data File Date Data Files Report Difference

97380100 03/09/2002 104,470 104,470 0 97380200 02/23/2002 64,788 64,788 0 97380300 02/23/2002 35,334 35,334 0 97380400 02/23/2002 50,565 50,565 0 97380500 03/09/2002 185,709 185,710 1 97380600 03/09/2002 81,592 81,592 0 97380800 02/23/2002 65,169 65,169 0 97380900 03/09/2002 30,996 30,996 0

Totals 618,623 618,624 1

The counts in the payroll data files differed from the headcount reports by one employee, which is under the headcount reporting threshold of 2 percent. Procedure. 6. Calculate employer and employee contributions for retirement, health benefits, and life insurance. 6.a. Calculate retirement withholdings and contributions for the four pay periods selected. 6.a.i. Multiply the CSRS and FERS payroll base by the withholding and employer contribution rates required by law. 6.a.ii. Compare the calculated totals with related amounts shown on the RITS submissions. Report any variances (e.g., gross rather than net) between the calculated amounts and the amounts reported on the RITS submissions greater than 5 percent of the amounts on the RITS submission. Auditor Action. We calculated the total CSRS and FERS retirement employee withholdings and employer contributions for the pay periods ended: February 23, 2002, for the four payroll entities 0200, 0300, 0400, and 0800; March 9, 2002, for the four payroll entities 0100, 0500, 0600, and 0900. Employee withholding rates for CSRS and FERS were 7.00% and 0.8%, respectively. Employer contribution rates for CSRS and FERS were 8.51% and 10.7% respectively. The differences between the calculated total of CSRS and FERS employee retirement withholdings and employer contributions and the related amounts shown on the RITS submission did not exceed the agreed-upon threshold of 5 percent for any of the 8 payroll data files. Differences for CSRS and FERS employee withholdings varied between 0.02 percent to 0.10 percent and 0.00 percent to 0.27 percent respectively, as shown in Table 5 and Table 6.

14

Table 5 shows the comparison of DFAS records of CSRS and FERS employee withholdings to amounts withheld per RITS, and the percent difference.

Table 5. Comparison of CSRS and FERS Employee Withholding per DFAS to RITS

CSRS CSRS FERS FERS Calculated Per DFAS Calculated Per DFAS

Data Employee Employee Percent Employee Employee Percent File Withholding Withholding Difference Withholding Withholding Difference 100 5,630,031 5,633,571 -0.06% 794,035 794,048 0.00% 200 3,575,578 3,578,562 -0.08% 455,071 455,118 -0.01% 300 1,615,756 1,616,060 -0.02% 166,325 166,779 -0.27% 400 2,537,370 2,538,807 -0.06% 320,943 320,990 -0.01% 500 10,442,885 10,448,938 -0.06% 1,568,136 1,568,168 0.00% 600 4,944,699 4,945,740 -0.02% 542,826 543,973 -0.21% 800 4,170,341 4,169,019 0.03% 472,039 472,020 0.00% 900 1,409,973 1,411,395 -0.10% 187,738 187,741 0.00%

Totals 34,326,633 34,342,092 -0.05% 4,507,113 4,508,837 -0.04% Table 6 shows the comparison of DFAS records of CSRS and FERS employer contributions to amounts withheld per RITS, and the percent difference.

Table 6. Comparison of CSRS and FERS Employer Contribution per DFAS to RITS

CSRS CSRS FERS FERS Calculated Per DFAS Calculated Per DFAS

Data Employer Employer Percent Employer Employer Percent File Contribution Contribution Difference Contribution Contribution Difference 100 7,217,950 7,216,520 0.02% 10,620,216 10,620,215 0.00% 200 4,579,699 4,580,406 -0.02% 6,086,580 6,086,407 0.00% 300 2,088,882 2,088,699 0.01% 2,224,595 2,226,785 -0.10% 400 3,250,936 3,252,122 -0.04% 4,292,613 4,292,603 0.00% 500 13,392,089 13,396,830 -0.04% 20,973,813 20,975,360 -0.01% 600 6,292,728 6,293,166 -0.01% 7,260,295 7,260,239 0.00% 800 5,368,592 5,366,122 0.05% 6,313,527 6,313,526 0.00% 900 1,798,580 1,799,585 -0.06% 2,510,999 2,511,002 0.00%

Totals 43,989,456 43,993,450 -0.01% 60,282,638 60,286,137 -0.01% Procedure. 6.b. Calculate employee withholdings and employer contributions for health benefits for the four pay periods selected. Auditor Action. We obtained the number of employees enrolled in each health insurance plan for each payroll data file from data provided by DFAS as RITS submissions. We obtained the official subscription rates for health benefits issued by OPM for all plans and options available to Federal employees from the OPM website.

15

We extended and added the totals and compared the results with the health insurance withholding and contribution amounts shown on the OPM Collection and Deposit System Standard Form 2812. All but one Payroll Office was below the 5 percent reporting threshold. Payroll Office 900 had a variance of 5.26 percent for employee contributions for the February 23, 2002, pay period. We would also like to note that for the pay period ended January 26, 2002, the overall difference for employee contributions was 4.84 percent. These differences are the highest that we have seen since the inception of the agreed-upon procedures. Management Comments. DFAS nonconcurred and provided additional information. DFAS stated that the discrepancy described for database 380900 is significantly overstated. DFAS stated that the difference in the SF 2812 amount and the extract report is only a 0.0037 percent difference. DFAS also commented that it did not provide employer contribution amounts on the extract report because they were not included in the request from the Inspector General of the Department of Defense (IG DoD). DFAS stated that the auditors attempted to extrapolate the employer contribution amounts by using the number of employees in each health code for the pay period examined. DFAS asserted that this method is not acceptable for determining the proper health benefits amount because it would not include any retroactive actions. DFAS also commented that this method would not include proper amounts for part-time employees, they are calculated based on scheduled hours, as well as OMB rates. Audit Response. The management comments are responsive. We used the methods recommended in the agreed-upon procedures to recalculate health benefits. Step 6.b.i requires multiplying the number of employees enrolled in each health benefits plan and plan option by the employee withholdings and employer contributions for the plan and option. Step 6.b.ii requires summing the totals in Step 6.b.i. and comparing the result with the health benefit withholding and contribution amounts shown on the SF 2812. Step 6.b.ii. requires us to report any variances (e.g., gross rather than net) between the calculation amounts and the amounts reported on the SF 2812s greater than 5 percent. We are reluctant to request any additional data from DFAS due to our prior experience of extracting accurate and timely information from its database. DFAS does not have an accounting system to record payroll data, but instead relies on the Defense Civilian Pay System database that takes sometimes up to 8 weeks to get the data. If the recalculations exceed the threshold for reporting a variance, we will request retroactive actions and data on part-time employees for further explanation of the difference. Procedure. 6.c. Calculate the Basic Life Insurance employee withholdings and employer contributions for the four pay periods selected. Auditor Action. We obtained the number of employees enrolled in each life insurance plan for each payroll data file from data provided by DFAS as RITS submissions. We obtained the cost of employee withholdings and employer premiums for life insurance from the Access database provided by DFAS. We extended and added the totals and compared the result with the life insurance withholding and contribution amounts shown on the OPM Collection and Deposit System Standard Form 2812. All Payroll Offices are below the 5 percent reporting threshold.

16

Procedure. 6.d. Calculate the Option A and Option C Life Insurance coverage withholdings for the four pay periods selected by using detail payroll reports used to reconcile the RITS reports in Step 1. Auditor Action. We obtained the number of participating employees from DFAS for each payroll data file. We totaled the individual withholding for Option A and Option C for each payroll data file and each date. For the pay period ended October 20, 2001, the overall difference for Option A was 0.13 percent. For the pay period ended January 26, 2002, the overall difference for Option A was 0.10 percent. For the pay period ended February 23, 2002, the overall difference for Option A was 0.13 percent. For the pay period ended March 9, 2002, the overall difference for Option A was 0.09 percent. The differences between the amounts reported to OPM by DFAS and the amounts we totaled were significantly less than the reporting threshold of 2 percent in the agreed-upon procedures for Option A. For the pay period ended October 20, 2001, the overall difference for Option C was 0.17 percent. For the pay period ended January 26, 2002, the overall difference for Option C was 0.27 percent. For the pay period ended February 23, 2002, the overall difference for Option C was 0.18 percent. For the pay period ended March 9, 2002, the overall difference for Option C was 0.12 percent. The differences between the amounts reported to OPM by DFAS and the amounts we footed were significantly less than the reporting threshold of 2 percent in the agreed-upon procedures for Option C.

17

Appendix A. Scope and Methodology

We reviewed a sample of the data and documentation supporting the $2.3 billion in payroll withholding reported each year by DFAS to OPM for DoD civilian personnel. The total annual payroll for about 675,000 DoD civilian employees is $25.2 billion. We selected for review the payroll files and supporting documentation for the payroll periods that ended October 20, 2001; January 26, 2002; February 23, 2002; and March 9, 2002. The DFAS payroll history database did not support the amount of DoD payroll withholding reported to OPM since the detailed data provided to us for the audit, when added together, did not agree with amounts reported to OPM. We made recommendations addressing this deficiency for FY 2000 in Inspector General Department of Defense Report No. D-2001-109, issued April 27, 2001. We had previously made recommendations addressing this deficiency for FY 1999 in IG DoD Report No. D-2000-156, issued June 29, 2000. DFAS had offered an alternative procedure, referred to as “data warehousing,” during mediation to maintain data for 6 years when fully implemented. IG DoD considers this an acceptable alternative because it supports U.S. Treasury record retention requirements.

We reviewed data and documentation supporting gross pay and payroll withholdings that DFAS reported to OPM for the four pay periods ended October 20, 2001; January 26, 2002; February 23, 2002; and March 9, 2002. We also reviewed management controls over the reporting process. We compared the payroll data files with employee personnel forms for 280 randomly selected employees for gross pay, retirement, health insurance, and life insurance.

We performed the agreed-upon procedures required by OMB, including verification of the payroll data file totals and recalculations of insurance and retirement withholdings.

We did not perform an audit but performed additional procedures based on auditing standards from February 2002 through August 2002. .

Use of Computer-Processed Data. We did not evaluate the general and application controls of the Defense Civilian Pay System that processes payroll data, although we did rely on data produced by that system to conduct the audit. We did not evaluate the controls because we determined data reliability by totaling the data provided to us from the system and comparing the totals to summary documents previously prepared from data in the system. Not evaluating the controls did not affect the results of the audit.

General Accounting Office High-Risk Area. The General Accounting Office has identified several high-risk areas in DoD. This report provides coverage of the Defense Financial Management high-risk area.

Management Control Program Review

DoD Directive 5010.38, “Management Control (MC) Program,” August 26, 1996, and DoD Instruction 5010.40, “Management Control (MC) Program Procedures,”

18

August 28, 1996, require DoD managers to implement a comprehensive system of management controls that provides reasonable assurance that programs are operating as intended and to evaluate the adequacy of the controls.

Scope of Review of the Management Control Program. We reviewed the adequacy of DoD personnel offices’ management controls over official personnel files. Specifically, we reviewed DoD personnel offices’ management controls over accuracy of personnel elections for payroll withholding, transmission of payroll withholding data to DFAS, and retention of personnel payroll withholding election data in the official civilian personnel files. We reviewed the annual statements of assurance by the Military Departments and Defense agencies to determine whether they disclosed the inconsistency between official personnel files and Defense Civilian Pay System payroll withholding data.

Adequacy of Management Controls. We identified a management control weakness for DoD personnel offices as defined by DoD Instruction 5010.40. DoD personnel offices’ management controls for accuracy of personnel payroll withholding elections, timely transmission of personnel payroll withholding data to DFAS, and retention of personnel payroll withholding elections in official personnel files were not adequate to ensure that civilian personnel were paid and withholdings made properly, that civilian personnel payroll withholding data were transmitted in a timely manner, and that documents and data supporting payroll withholding were retained in the official personnel files. We previously reported this management control weakness in Inspector General Report No. D-2002-070, issued March 25, 2002. Recommendations 1.a., 1.b., and 2. in that report have been implemented and will improve DoD personnel office payroll withholding procedures. A copy of that report was provided to the senior official responsible for management controls of the personnel offices of the Military Departments and Defense agencies.

Adequacy of Management’s Self-Evaluation. Self-evaluation by the Military Departments and Defense agencies did not identify the weakness because management did not identify the area as an assessable unit.

Prior Coverage

The IG DoD and the Air Force Audit Agency have conducted multiple reviews related to civilian payroll information, controls over the payroll process, and payroll expenses. Unrestricted IG DoD reports can be accessed over the Internet at www.dodig.osd.mil/audit/reports. Unrestricted Air Force Audit Agency reports can be accessed over the Internet at www.afaa.hq.af.mil.

19

Inspector General of the Department of Defense

IG DoD Report No. D-2002-070, “DoD Payroll Withholding Data for FY 2000,” March 25, 2002

IG DoD Report No. D-2001-109, “DoD Payroll Withholding Data for FY 2000,” April 27, 2001

IG DoD Report No. D-2000-156, “DoD Payroll Withholding Data for FY 1999,” June 29, 2000

20

Appendix B. Report Distribution

Office of the Secretary of Defense Under Secretary of Defense (Comptroller)/Chief Financial Officer

Deputy Chief Financial Officer Deputy Comptroller (Program/Budget)

Under Secretary of Defense for Personnel and Readiness

Department of the Army Assistant Secretary of the Army (Financial Management and Comptroller) Assistant Secretary of the Army (Manpower and Reserve Affairs) Auditor General, Department of the Army

Department of the Navy Assistant Secretary of the Navy (Manpower and Reserve Affairs) Naval Inspector General Auditor General, Department of the Navy

Department of the Air Force Assistant Secretary of the Air Force (Financial Management and Comptroller) Assistant Secretary of the Air Force (Manpower, Reserve Affairs, Installations, and

Environment) Auditor General, Department of the Air Force Other Defense Organizations Director, Defense Commissary Agency Director, Defense Finance and Accounting Service Director, Defense Logistics Agency Director, Defense Information Systems Agency Director, Civilian Personnel Management Services

Non-Defense Federal Organizations and Individuals Office of Management and Budget Office of Personnel Management

21

Congressional Committees and Subcommittees, Chairman and Ranking Minority Member

Senate Committee on Appropriations Senate Subcommittee on Defense, Committee on Appropriations Senate Committee on Armed Services Senate Committee on Governmental Affairs House Committee on Appropriations House Subcommittee on Defense, Committee on Appropriations House Committee on Armed Services House Committee on Government Reform House Subcommittee on Government Efficiency, Financial Management, and

Intergovernmental Relations, Committee on Government Reform House Subcommittee on National Security, Veterans Affairs, and International Relations,

Committee on Government Reform House Subcommittee on Technology and Procurement Policy, Committee on

Government Reform

Defense Finance and Accounting Service Comments

Final Report Reference

23

Pages 3-4 Page 15

Final Report Reference

24

Team Members The Defense Financial Auditing Service, Office of the Assistant Inspector General for Auditing of the Department of Defense prepared this report. Personnel of the Office of the Inspector General of the Department of Defense who contributed to the report are listed below.

Paul J. Granetto Richard Bird David F. Vincent Thomas J. Winter Joseph A. Powell Jonathan M. Rabben Joseph Bilock Mark Heim Angel Beltran Ann Thompson